Our knowledge of how Arctic freshwater ecosystems will respond to continued climate change and variability is fundamentally limited by logistical difficulties of such remote research, resulting in relatively sparse long-term baseline data on these systems. This research applies a unique paired-watershed approach (i.e., two similar, adjacent lakes and catchments) to help address these limitations, which provided an opportunity to identify how broad-scale factors are filtered or modified by site-specific characteristics. My first main objective was to document the seasonal hydrochemical variability of runoff and influences on lake chemistry. Both lakes appear to be relatively insensitive to seasonal hydroclimatic variability, largely because periods of high discharge were also characterized by lower concentrations of dissolved and particulate matter, but also because of the relatively long lake water turnover time that suggests the effects of climatic and environmental changes would be felt later in these systems than in lakes and ponds with smaller lake volumes. My second objective was to investigate spatial and temporal trends in the lake diatom communities in order to refine subsequent paleoenvironmental reconstructions. A critical aspect of this objective was testing how faithfully the whole lake diatom community was represented in deep lake surface sediments where sediment cores are routinely collected. Most differences between the two lakes were largely accounted for with micro-environmental conditions associated with the specific sampling location. Also, both lakes exhibited a degree of disconnection between littoral benthic and profundal zones that manifested as an under-representation of the benthic community in deep lake surface sediments, with implications for paleoenvironmental interpretations. Finally, I present a multi-proxy record of environmental conditions in adjacent lakes spanning the period from pre-industrial times. Biogeochemical records reflected major changes in lake primary productivity and terrestrial organic matter accumulation beginning prior to 1950 in both lakes, pointing to profound environmental changes that culminated with the establishment of an appreciable diatom community in both lakes in the 1980s. Differences in the timing of changes between the two lakes point to differing threshold capacities to external forcings, and suggest that East Lake’s response to post-industrial climate change is advanced compared to West Lake. / Thesis (Ph.D, Geography) -- Queen's University, 2011-10-26 09:34:01.144

Three watersheds located in Westmoreland County, Virginia were monitored to evaluate effectiveness of forestry best management practices (BMPs) for minimizing hydrologic and water quality impacts associated with timber harvesting. One watershed was clear-cut without implementation of BMPs, one watershed was clear-cut with the implementation of BMPs and the third watershed was left undisturbed as a control. The 27 months of pre-harvest monitoring data and 17 months of post-harvest monitoring data were compared using the paired watershed regression analysis and the minimum detectable change (MDC) statistic. Analysis of the hydrologic data showed that peak discharge rates were not impacted by harvesting, regardless of whether BMPs were implemented. Harvesting with or without the implementation of BMPs resulted in no statistically significant change in stormflow volume. BMP implementation was found to be effective in minimizing harvesting impacts on stormflow total suspended solids (TSS), ammonia, total Kjeldahl nitrogen (TKN), total nitrogen (TN), sediment-bound nitrogen and total phosphorous (TP) concentrations and total flow TSS and TP concentrations. BMP implementation was also effective in minimizing harvesting impacts on stormflow TSS, TKN, sediment-bound nitrogen, TP and sediment-bound phosphorous loadings and total ammonia loading. The BMPs were not effective in reducing nitrate concentrations and loadings. The BROOK90 forest hydrology model was used to predict the site hydrology. Model predictions compared favorably with streamflow measurements from watersheds QN3 and QN4 until the time of harvest, but did not predict streamflows as accurately for the control watershed, QN5. / Master of Science